This is a first resubmission of our application. Our proposal has been revised extensively in response to the reviewer's comments. [unreadable] Some antipsychotic medications appear to increase the risk of diabetes. However, other evidence raises the possibility that schizophrenia is associated with an increased risk of diabetes in the absence of antipsychotic use. An important recent study found increased insulin resistance in drug-naive, newly diagnosed patients with schizophrenia compared to controls, but the same study also found much higher cortisol concentrations in the schizophrenia subjects than in the controls. As hypercortisolemia can cause insulin resistance, increased glucose concentrations in this and other studies may have been due to hypercortisolemia. Other confounding factors also may have operated in previous research. [unreadable] We propose to conduct 2 studies. 1) We will administer a glucose tolerance test to 4 groups matched on gender, age, body mass index, and certain lifestyle variables: drug-naive, recently diagnosed schizophrenia subjects; subjects with depression; subjects with adjustment disorder; and control subjects. We will test the hypothesis that schizophrenia group increased insulin resistance compared to the control group; inclusion of the other 2 groups will enable us to assess the role hypercortisolemia. We will also conduct exploratory analyses of inflammatory markers C-reactive protein and interleukin-6, which are independent markers of diabetes risk. 2) In an open trial in drug-naive newly diagnosed subjects with psychosis, we will test the hypothesis that greater insulin resistance prior to treatment predicts a larger increase in insulin resistance with olanzapine treatment. These studies will assess whether schizophrenia is associated with impaired glucose tolerance, and whether this abnormality predicts the subsequent risk of impaired fasting glucose with olanzapine administration.